Classifications

• • H—ELECTRICITY
  • H04—ELECTRIC COMMUNICATION TECHNIQUE
  • H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
  • H04N9/00—Details of colour television systems
  • H04N9/77—Circuits for processing the brightness signal and the chrominance signal relative to each other, e.g. adjusting the phase of the brightness signal relative to the colour signal, correcting differential gain or differential phase
• • H—ELECTRICITY
  • H04—ELECTRIC COMMUNICATION TECHNIQUE
  • H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
  • H04N9/00—Details of colour television systems
  • H04N9/64—Circuits for processing colour signals
  • H04N9/68—Circuits for processing colour signals for controlling the amplitude of colour signals, e.g. automatic chroma control circuits

Definitions

• the present invention relates to controlling the contrast in a picture signal.
• a video signal for use as input to an electronic video display consists of a number of sub signals, each sub signal representing color information for different areas of the picture, and the color information is represented by basic colors, i.e. RGB (red, blue, green) or YUV (luminance & chrominance).
• a problem with contrast amplification is that there is a contrast limit corresponding to the electronic video display or the D/A converter in the signal processor, meaning that the electronic video display is adapted to receive a limited range of input signals and the D/A converter in the signal processor is adapted to deliver a limited range of output signals.
• US 5,349,390 (Attorneys' docket PHN 14.195).
• US 5,349,390 describes a device for reducing the amplitude range, which includes a first contrast reduction device receiving a picture signal for providing a contrast reduced picture signal; the contrast reduction device only reduces the contrast of the picture signal over the entire area when the picture signal exceeds a given first threshold (e.g. 90%) for relatively large areas.
• a given first threshold e.g. 90%
• the picture signal processor also includes a second contrast reduction device, which supplements the first contrast reduction device and which provides the contrast reduced picture signal by immediately reducing the contrast of a too bright part of the first contrast reduced picture signal as soon and as long as an instantaneous amplitude of the first contrast-reduced picture signal exceeds a given second threshold (100%).
• a second contrast reduction device which supplements the first contrast reduction device and which provides the contrast reduced picture signal by immediately reducing the contrast of a too bright part of the first contrast reduced picture signal as soon and as long as an instantaneous amplitude of the first contrast-reduced picture signal exceeds a given second threshold (100%).
• This implementation is built with analog circuitry and is based on a feedback loop. Further an analog implementation is very cumbersome and noise sensitive, and it is hard to make adjustments for an optimum response. Further the picture might lose details in the bright areas because of the clipping of the small peaks.
• a method is obtained where it is possible to overdrive the picture signal and to obtain maximal light output from the display.
• the contrast may also be referred to as the amplitudes of the basic colors and the basic colors could e.g. be RGB, YUV.
• a video display device may be a computer monitor, television, projector etc.
• the method may easily be implemented by digital circuits, whereby the advantages associated with digital circuits may be obtained.
• the step of splitting the sub signal into two groups of picture information may comprise the step of splitting the selected sub signal into at least two groups of color information comprises the step of transforming the selected sub signal from being represented by the first set of basic colors to a second set of basic colors - e.g. U, V and Y.
• the splitting is a very easy task since UV represents the chrominance and Y only represents luminance.
• the first value may be determined by multiplying the reciprocal of the amplitude of that basic color having the highest amplitude and being above the predefined value with the predefined value, and wherein the reduction of saturation is performed by multiplying the amplitudes of the signals of the first group of sub signals with the first value.
• the second value may be determined by multiplying the reciprocal of the amplitude of that basic color having the highest amplitude and being above the predefined value with the predefined value, and wherein the reduction of all basic colors is performed by multiplying the amplitudes of the signals of both the first and second group of sub signals with the second value.
• the first and/or said second value may decrease from 1 to a limit value.
• the limit value may be determined by multiplying the reciprocal of the amplitude value of the basic color having the highest amplitude with said predefined value.
• the reciprocal of the amplitude value of the basic color having the highest amplitude may be defined in lookup tables.
• the first and second values may be defined in lookup tables.
• the lookup table may be implemented as a combination of at least one sparse table and at least one linear interpolator.
• the steps of performing the reduction of saturation and performing the reduction of all basic colors may be performed in a small insignificant part of the picture in which at least one amplitude of the basic colors exceeds the predefined value. These small areas could i.e. be subtitles. Contrast reduction of a small insignificant part may be combined with contrast reduction of a large significant part of the picture in which at least one amplitude of the basic colors exceeds the predefined value. These large areas could i.e. be faces.
• the means recited in the signal processing device claim may be computer means, such as a CPU.
• the CPU could be customized, but it could also be a more general pre-programmed CPU.
• Other examples could be a digital signal processor (DSP) or other digital logic.
• DSP digital signal processor
• Fig. 1 shows a block diagram of an embodiment of the clipper for small areas of the present invention
• Fig. 2 shows the embodiment of the block diagram in Fig. 1, where lookup tables are introduced
• Fig. 3 shows a display apparatus comprising a display and a processing device
• Fig. 4A-B show clipping characteristics varying from very soft to hard
• Fig. 4 A shows the output signal as a function of the largest input signal
• Fig. 4B shows the multiplication factor as a function of the largest input signal
• Fig. 5 shows a block diagram of an embodiment of the detector for large area overdrive conditions of the present invention
• Fig. 6 shows the RGB values as a function of the maximum of the RGB values according to another version of a clipper
• Fig. 7 shows the RGB values as a function of the maximum of the RGB values according to another clipper, where the color changes from black to blue in a linear ramp.
• an advantageous method of controlling contrast in at least a group of sub signals in a picture is by consecutively following the steps of:
• a third step could be introduced as a supplement to the above steps. This step reduces the contrast of the entire picture, if the picture contains large bright areas, such as faces. Thereby clipping is avoided in large areas, preventing especially loss of detail in facial texture. In an alternative embodiment these larger bright parts or even the total picture could be contrast reduced by the above steps.
• the input sub signal is represented by the YUV color space, and these signals can be changed in contrast or brightness.
• Change of contrast multiplies YUV with a gain factor (con) and change of brightness adds an offset term (bri) to Y. This change in contrast and brightness is the main reason why the output signal may be too large and clipping of the (white) peaks may be necessary.
• the YUV colors are converted to RGB according to the following conversion matrix:
• the above form separates the contributions from the luminance channel (Y) and the chrominance channel (U&V) to the RGB color space. Both channels are related to contrast, but only the chrominance channels relate to saturation, and therefore it is now possible to control the saturation separately by first adjusting the amplitude of the contribution from the chrominance channel. Therefore the RGB contribution to Y and UV can be separated as shown below:
• the total RGB signal is obtained by adding R_Y to R UV, G_Y to GJJV and
• a first overdrive factor is determined. This first overdrive factor is determined according to a maximum value for each basic color RGB; i.e. the value could be set according to a limit, such as a maximum contrast value to be used in the video display without overdriving the display, or according to an input limit to the video display specified by the manufacturer of the display adapter.
• the amplitude of all the basic colors in the sub signal(s) is reduced according to the present invention by reducing only the U and V part or the saturation part of the sub signal(s) with the first overdrive factor (by multiplying the basic colors with the overdrive factor 1/x).
• the amplitude of the smallest signal(s) of the RGB sub signals may be increased, but this is not a problem.
• the total RGB signal is calculated again by adding the Y part and the reduced UV part of the RGB signal as stated above.
• the total RGB signal is then reduced with a second overdrive factor. Again it is determined if the amplitude of one of the RGB sub signals is above a predefined value. If this is the case, then they are reduced with a second overdrive factor. Again the overdrive factor is determined according to the maximum value for each basic color RGB; i.e. the value could be set according to a limit, such as a maximum contrast value to be used in the video display without overdriving the display, or according to an input limit to the video display specified by the manufacturer of the display adapter. If one of the RGB sub signals is above the predefined value, then the amplitude of all the signals is reduced according to the present invention by reducing all the YUV parts or the contrast part of the RGB sub signals with the second overdrive factor.
• the total RGB signal is calculated again by adding the reduced YUV parts of the RGB signal as stated above.
• RGB may be no larger than +1.000, then they have to be reduced again by a factor of 1.50* .
• our overdrive factor is 1.50.
• Fig. 2 the embodiment of Fig. 1 is shown where soft clipping is introduced using lookup tables (LUT).
• LUT lookup tables
• different gain values are stored and are chosen according to the amplitude value of the maximum color in the set of basic colors to be reduced.
• a digital time-discrete soft-clipper as mentioned above necessarily works in feed-forward fashion by: Taking the maximum of the 3 input signals (usually RGB),
• Multiplying each of the 3 inputs with the same gain (also referred to as 100% coupling).
• 100 % coupling means that the ratio of the 3 input signals (R:G:B or Y:U:V) is not disturbed, so tint and (if you so choose) saturation are preserved. This is achieved by multiplying (R,G,B) or (Y,UN) with the same factor.
• the content of the lookup table determines the nature of the clipping by holding information, such as:
• Fig. 3 shows a display apparatus 1 according to the present invention comprising a processing device 3 and a display 5.
• the display apparatus could be a television, monitor etc.
• Fig. 4A-B show clipping characteristics varying from very soft to hard
• Fig. 4A shows the output signal as a function of the largest input signal
• Fig. 4B shows the multiplication factor as a function of the largest input signal.
• the figure shows a graph of the input-output transfer function of the clipper, while varying the hardness of the clipping.
• Hard-clipping is typically allowed only on small peaks, such as subtitles. In the presence of large bright areas, particularly faces, it is then tuned to softer clipping, and/or the global contrast value must be reduced.
• Fig. 4A with soft clipping we start earlier with reducing the gain, even before one of the R,G,B-signals is overdriven; and as the signal is overdriven more, we will reduce the gain some more.
• Fig. 2 LUT UV and LUT RGB.
• the differential gain being the variation of the output signal divided by the variation of the input signal, is not suddenly reduced to zero. Instead we are reducing the gain earlier than necessary. Only at hard overdrive do we reach the maximum output level, we are sacrificing a bit of light output to make a more natural image. We are preserving some details in bright areas.
• the saturation reduction in the first stage makes the highest of (R,G,B) lower, and the others higher. This means that less contrast reduction needs to be made at the second stage.
• the end result is fewer artifacts due to hard contrast clipping, as well as getting a bit more brightness on the screen.
• the 2 lookup tables are not implemented as full tables. Instead they each use a sparse table and a linear interpolator. This saves on memory size while not essentially reducing the flexibility.
• An example in C code of an implementation of a lookup table for the contrast reduction is shown below:
• Input to the lookup table is an 11 -bit number M. This is split into a 6-bit MSB part idx and a 5-bit LSB part rem.
• minus is used in the interpolation in the interpolator because the content of the clipping table is always decreasing, and then only non-negative numbers for delta are stored.
• the factor F is determined from input signals R,G,B and applied to output signals R,G,B, so there is not a feedback path as in US 5,349,390, where the factor F would be determined from the output signals R,G,B.
• the present invention introduces a "faces detector". It measures how large a fraction of the output signal is in the upper regions of the signal range, where the soft-clipper may be set to a low differential gain. This is an indication of loss of details in large areas, e.g. facial texture.
• the way it functions is by means of a reverse peak detector: it resets quickly on dark pixels (between the subtitles) and it follows bright pixels slowly. They can only be counted, if a sufficient number of consecutive pixels are bright then.
• the speed with which the filter follows consecutive white is programmable: the maximum up-going slope of an integrator is set.
• Fig. 5 shows a block diagram of an embodiment of the detector for large area overdrive conditions of the present invention, this embodiment is based on non-linear filtering.
• the peak-drive measurement is presented as a small (8-bins) histogram. This gives more relevant information than a single peak-drive value.
• the bins' borders are at V 2 ,
• Contrast reduction and desaturation can also be combined as explained in the following, which is another embodiment of the invention.
• Fig. 6 we see values of the sub signals R, G, B as functions of the maximum contrast value, for simplicity 1 is the maximum contrast value for R, G, B in this example.
• the sub signals is below clipping level (x ⁇ 1), a processed signal going to the electronic video display is equal to the sub signals. If x > 3, meaning that at least one of the sub signals is over three times too high to remain in the non-clipping region, the output will be a constant "peak white" of which we can choose the exact composition of R, G, B.
• the output is a mix (i.e.
• the processing shown in Fig. 6 could be carried out in the "linear-light domain” or alternatively in the "gamma-corrected signal domain".
• the signals correspond linearly to the luminance output (measured in cd/m 2 ).
• the signals correspond to luminance 0-45 approximately.
• the factor 1 /(xwhite- 1) is implemented with a lookup table.
• An advantage of using a lookup table for xwhite is that we can easily choose xwhite very large for extremely saturated colors (lowest of (rin, gin, bin) close to zero). In that case the saturated colors will be prohibited to desaturate, even if their input strength (x) gets high.

Priority Applications (1)

Applications Claiming Priority (4)

Publications (1)

Family

ID=8180137

Family Applications (1)

Country Status (6)

Families Citing this family (21)

Family Cites Families (23)

• 2002
  • 2002-04-02 KR KR1020027016808A patent/KR100911815B1/ko not_active IP Right Cessation
  • 2002-04-02 WO PCT/IB2002/001095 patent/WO2002085037A1/en active Application Filing
  • 2002-04-02 CN CNB028011260A patent/CN1305319C/zh not_active Expired - Fee Related
  • 2002-04-02 JP JP2002582633A patent/JP4130366B2/ja not_active Expired - Fee Related
  • 2002-04-02 EP EP02713160A patent/EP1380177A1/de not_active Ceased
  • 2002-04-08 US US10/117,827 patent/US7050114B2/en not_active Expired - Lifetime

Non-Patent Citations (1)

Also Published As

Similar Documents

Legal Events